Once touted as the cleaner alternative to gasoline due to lower CO2 emissions, the use of diesel as a transportation fuel is under intense scrutiny following Volkswagen’s scandal in 2015. Since then, academics and media outlets have publicized the adverse effects of NOx emissions on air quality and public health. In a somewhat knee-jerk reaction, many governments around the world called for an outright ban of diesel vehicles.
Lux Research compiled a non-exhaustive list of cities around the world that announced intentions to ban diesel vehicles. While some cities called for a blanket ban, others are introducing restrictions to limit the number of diesel vehicles, a step we believe will eventually move towards a ban.
Just a few weeks ago, Tesla announced the release of the official beta-version of its Autopilot. Only a few months off schedule from the original plans for an end-of-summer deployment, the system was immediately met with huge acclaim from Tesla owners. Many of those same owners started immediately abusing the system, using it in scenarios it wasn’t intended for and pushing the technology to its limits.
Automakers, in turn, have expressed concern – if not outright disdain – for what Tesla is doing. The concerns of automakers, many of which were recently voiced at the Tokyo Auto Show, is that if an accident happens with one of these vehicles, there will be an immediate backlash on the regulatory environment and progress of all autonomous vehicle technology.
What We Think
Automakers have every right to be concerned about Tesla’s latest product release. The company is essentially operating in a legal loophole, where it does not consider its technology to be “autonomous” but rather an advanced driver safety feature. As a result, it’s pushed its product to market while operating in this legal grey zone between autonomous and autopilot. Back in 2014, when California required Google to install steering wheels and pedals back into its driverless cars (client registration required), automakers began debating the language of autonomous versus autopilot. Attempting to distinguish between the two is somewhat ridiculous, and seems like a ploy from automakers like Tesla to ensure that the blame for any accident falls to the driver rather than the vehicle. Ideally, recent announcements like those from Volvo (client registration required) will help clear up some of this confusion, but in the end it will be up to regulators to decide what’s legal and where liability falls.
In the meantime, Tesla is focusing on the good news: in the three weeks since deployment, there have been lots of near misses but no major accidents. The system even appears to be learning and improving. One driver even claimed that the autopilot saved the vehicle from an otherwise unavoidable accident. However, for all the positive press, it’s important to note that the recent accident avoidance by the Tesla Uber actually highlights the confusion over advanced driver-assist systems and how they work.
The driver says Autopilot deserves credit for avoiding the accident; however, it was really the car’s forward-collision warning with automatic braking that avoided the accident. Forward collision warning and braking is increasingly a standard safety feature among many vehicle brands, not just Tesla, and is technically a component of the Autopilot but not actually its distinguishing feature. Rather, it’s the system’s ability to change lanes that makes it a significant feature jump from previous advanced driver assist systems (ADAS).
The problem with Autopilot style systems is that you cannot trust drivers to take control of the car, particularly once they become comfortable with the technology. It will lead to complacency and even longer re-engagement times (client registration required). What OEMs need to focus on is how to predict the scenarios in which the car can no longer self-drive and pass control to the driver with more than sufficient time – it cannot be a “just-in-time” operation, as that will almost certainly result in accidents. Today, the focus on partial autonomy on the highway is driven by the fact that the highway is actually really predictable in terms of traffic patterns, other vehicle behaviors, and accurate/adequate maps. Most OEM’s are building for the situation of “full autonomy,” but confined to the highway environment and passing control as an exit, comes situations where the car can give ample warning.
What many of those automakers are concurrently working on is how to track driver engagement – either through facial recognition or eye tracking – in order to enable a safe passing of controls. Driver tracking is something that Tesla has decided to entirely forgo, instead opting to simply warn its drivers that they are still responsible for their vehicles. A lack of driver tracking, combined with the fact that Tesla’s Autopilot currently lacks a safety system that would prevent the driver from turning it on in inappropriate situations – such as urban streets – could spell disaster for the company. So far, Tesla’s Autopilot has performed relatively admirably, and thankfully there have been no major issues, but there is still plenty of reason to be concerned.
The Insurance Institute for Highway Safety (IIHS) has announced it will now test how well cars can autonomously avoid crashes. For a car to earn the IIHS’ “Top Safety Pick+,” it will have to have at least a basic level of frontal crash prevention. Meanwhile, the European New Car Assessment Programme (NCAP) is also now testing vehicles for crash avoidance, and will start including the test results in its 2014 car ratings.
What We Think
Crash prevention is one of the key milestones for car makers on the road to fully autonomous vehicles – specifically, level three of five (client registration required). However, thus far OEMs have deployed crash avoidance systems with vastly differing capabilities, potentially confusing consumers and slowing adoption: IIHS tested 74 vehicles offering some variant of this technology today, but found that only seven – from Cadillac, Mercedes-Benz, Subaru, and Volvo – can actually avoid crashes and significantly slow down at 25 mph.
The IIHS and NCAP crash avoidance testing effectively provides a useful benchmark with which consumers can evaluate OEMs’ first vehicle autonomy efforts. By tying this crash avoidance test to its “Top Safety Pick+” award, the IIHS is now providing an influential driving force for autonomous vehicles. Clients should expect that more consumer information agencies (like the U.S. NCAP) and regulatory bodies will start testing and dictating mild vehicle automation. However, full autonomy is still a decade away (client registration required).
The automobile is at a turning point, unprecedented in its 100+ years of history. Rising gas prices, stricter fuel economy standards, a progressively more environmentally conscious customer base and forward-looking business models are all breaking the traditional automotive ecosystem. In response, OEMs are evolving their partnership webs in order to endure and compete.
This week’s graphic comes from a recent Lux Research report in which analysts examined the growing web of cross‐cutting industry relationships to see how automakers compared on the Lux Partnership Grid. Companies were scored on two metrics, partnership strength and technology diversity. Based on these scores, each automaker fell into one of the four quadrants in the graphic above.
Lone Wolves represent OEMs that are continuing business as usual, and have little footprint in the emerging technologies that threaten the status quo. The Dilute ecosystem quadrant encompasses automakers with a few partnerships scattered across a variety of technologies. The Siloed ecosystem includes OEMs that are putting their faith in a few, or in some cases one technology, while the Expansive ecosystem hosts OEMs who average nearly 16 partnerships apiece, representing an average of eight unique technologies.
A review of the partnership grid reveals multiple trends:
Small companies tend to cluster in the Lone Wolves region, while large corporations partner ambitiously in a variety of areas so they group more in the Expansive quadrant. One exception is Fiat-Chrysler, which has only three, unsubstantial partnerships to its name.
Daimler, GM, and Toyota lead the pack. All have formed strong partnerships in pursuit of technical diversity, placing them squarely in the Expansive ecosystem. Daimler has developed partnerships that span multiple key emerging technologies, including a JV with Toray to make carbon fiber reinforced plastics (CFRP) (Client registration required), a JV with Evonik (Li-Tec) to make Li-ion batteries (Client registration required), and participation in Europe’s Clean Energy Partnership for establishing fuel cell vehicles and hydrogen fueling infrastructure.
Meanwhile, General Motors (GM) has emerged from a low point in its corporate history to emerge as a future looking company by partnering in a variety of technologies, and investing in companies at a variety of stages. It has invested in battery start-ups like Envia Systems (Client registration required) and Sakti3 (Client registration required), and in the fuels space with ethanol companies Mascoma (Client registration required) and Coskata (Client registration required). In the materials space, GM also sees value in CFRP, forming a JV with Japanese materials company Teijin.
Lastly, Toyota is leveraging its leadership in hybrids, to position itself for advances elsewhere, such as advanced electrification (via its investment in Tesla). It also retains activity in hydrogen powered fuel cell vehicles and infrastructure, where it teamed with Air Products and Shell to install the first pipeline-fed hydrogen station in the U.S. In materials, Toyota has partnered with Toray to source CFRP initially used for the hood and roof of one of its Lexus models. As with Daimler and GM, these activities will prepare Toyota to profit from an ever changing landscape where the vehicle of tomorrow may not look like anything conceived today, but will no doubt carry key technologies in the areas of energy storage, increased connectivity, and new materials.
The high and often volatile prices of platinum group metals (PGMs) has drawn many inventors and entrepreneurs to create technologies that reduce the need for PGMs in catalysts for the automotive, hydrocarbon processing and fuel cell industries. To assess which of these emerging catalyst developers stand to capitalize on this growing opportunity, Lux Research’s latest report ranks the field by technical value, business execution, and maturity and compares each company on the Lux Innovation Grid. Among the report’s key findings:
Nanostellar and SDCmaterials occupy the Dominant quadrant. SDCmaterials and Nanostellar recently moved into the Lux Innovation Grid’s Dominant quadrant, as the two strongest developers in the automotive catalyst space. Both firms boast strong technical solutions as well as solid partnerships – BASF and Volvo for SDCmaterials and several undisclosed European automakers for Nanostellar.
Headwaters Technology Innovation (HTI) retains High Potential, but may be stalled. HTI has lost momentum and missed a few milestones, but the firm can still claim commercial deployment of two catalyst products: a palladium catalyst that allows for simplified production of hydrogen peroxide (H2O2), and an oil refining catalyst that increases gasoline yield during catalytic reforming. However, HTI’s recent lack of significant developments and a parent company whose revenues have nearly been cut in half over the past three years all raise concerns that the company may soon see an even sharper drop in standing.
Once squarely in the Dominant quadrant, Catalytic Solutions may see trouble ahead. While Catalytic Solutions’ revenues are five times higher than any other company ranked in the report, they have dropped from $50.5 million in 2009 to $48.1 million in 2010. Meanwhile, net loss increased slightly, and the firm will likely be required to raise additional funds. The catalyst division also recently lost an automaker customer after its core technology failed to meet a required performance standard.